Transport in Gram-negative organisms is initiated by passage of the transported species through a transmembrane beta-barrel in the outer membrane. The transport of iron is particularly important for bacterial growth, and outer membrane iron transporters are therefore major vaccine targets against pathogens such as Neisseria, Haemophilus, and Yersinia. These transporters show high affinity and specificity for Fe(III)-ligand complexes, and require energy derived from the proton motive force across the inner membrane to transport ferric complexes. The required energy is provided by transient interaction with an integral inner membrane protein complex, TonB-ExbB-ExbD, resulting in a protein assembly that spans both the inner and outer membranes, as well as the periplasmic space. During 2004, we worked on the following projects: [1] Alternate mechanisms of ligand binding: structural evidence for iron-free citrate and ferric citrate binding to the TonB-dependent outer membrane transporter FecA: We recently solved the structures of FecA, the E. coli ferric citrate transporter, alone and in complex with dicitrate and diferric dicitrate, to resolutions of 2.1 A to 3.4 A. These three structures together shed new light on how TonB-dependent transporters recognize and bind their cognate ligands. We showed for the first time that FecA can bind both iron-free and ferric ligands, and that binding an iron-free ligand is a physiologically relevant phenomenon. The structure of FecA bound with iron-free dicitrate represents the first structure of a TonB-dependent transporter bound with an iron-free siderophore. We deduced the structural mechanism for discrimination between the iron-free and ferric siderophore: the binding of diferric dicitrate, but not iron-free dicitrate alone, causes major conformational rearrangements in the transporter. From these data we proposed a new model for ligand binding with implications for vaccine design: FecA binds iron-free dicitrate in the non-productive state or first step, followed by siderophore displacement or iron extraction to form the transport-competent, diferric dicitrate bound state in the second step. It is important to understand the complexities of ligand binding in order to develop vaccines or chemotherapeutic agents that target this family of transporters. Our discovery that FecA can bind an iron-free ligand led to a collaboration on outer membrane transporters that can bind iron-free ligands. [2] Plug domain folding and ability to bind ligand: the plug domain of a Neisserial TonB-dependent transporter retains structural integrity in the absence of its transmembrane beta-barrel: Neisseria meningitidis is the causative agent of bacterial meningitis. This blood-borne pathogen acquires iron from human transferrin (hTf = 80 kDa) through an outer membrane transporter complex, transferrin binding proteins A and B (TbpA = 100 kDa; TbpB = 68-85 kDa). TbpA and TbpB form a discrete complex to bind transferrin synergistically, yet each protein is also capable of binding transferrin on its own. TbpA is a TonB-dependent outer membrane iron transporter. We are interested in learning how this human pathogen can extract iron from transferrin and transport it into the periplasm. To evaluate the contribution of the plug domain to ligand recognition and binding, and to investigate its stability in the absence of the transmembrane beta-barrel, the N-terminal 160 residues (plug domain) of TbpA were overexpressed in both the periplasm and cytoplasm of E. coli. We found this domain to be soluble and monodisperse in solution, exhibiting secondary structure elements found in plug domains of structurally characterized TonB-dependent transporters, such as FepA and FecA. Although the TbpA plug domain is apparently correctly folded, we were not able to observe an interaction with human transferrin by isothermal titration calorimetry or nitrocellulose binding assays. This work contrasts with similar experiments performed on the plug domain of FepA, which was found to be unfolded but capable of binding ligand with a 100-fold reduced affinity. Our experiments suggest that the plug domain may fold independently of the beta-barrel, but extracellular loops of the beta-barrel are required for binding transferrin. [3] Maintenance of the cell envelope by outer membrane-associated proteins: structure of the OmpA-like domain of RmpM from Neisseria meningitidis: We recently solved the 1.9 A crystal structure of the C-terminal (periplasmic) domain of N. meningitidis RmpM, a protein that has been shown to interact with integral outer membrane proteins and that contains a domain responsible for non-covalent peptidoglycan binding. This domain is widely conserved across Gram-negative bacteria, with family members including the peptidoglycan-associated lipoprotein PAL and the flagellar motor protein MotB; ours is the first structure of such a domain. Although RmpM was predicted to be an outer membrane protein, our structure suggests that RmpM instead associates with integral outer membrane proteins through its N-terminal domain, while its C-terminal domain binds peptidoglycan. In this manner, RmpM can stabilize the cell envelope. From the structure of the OmpA-like domain of RmpM, we suggested a putative peptidoglycan binding site and identified residues that may be essential for binding. Both the crystal structure and solution experiments indicated that RmpM may exist as a dimer. This would promote more efficient peptidoglycan binding, by allowing RmpM to interact simultaneously with two glycan chains through its C-terminal, OmpA-like binding domain, while its (structurally uncharacterized) N-terminal domain could stabilize oligomers of porins and TonB-dependent transporters (such as TbpA) in the outer membrane.